1. Field of the Invention
The present invention relates to an information processing apparatus and a method of controlling the same, an information processing method, and a computer program.
2. Description of the Related Art
Recently, easy-to-use data generation apparatuses such as a digital camera, word processor, voice creating apparatus, and music creating apparatus have been developed. In addition, along with the progress of a computer having a large-capacity storage device or products incorporating a computer, cases that an enormous amount of data is stored in the storage device of a computer or a product incorporating a computer have greatly increased. It is therefore necessary to select desired data from the large amount of data stored in the storage device. However, selecting desired data while confirming the data one by one requires labor and time, resulting in large difficulty.
When desired data is text data, it is possible to some extent to select it by a character string search. For data having no text data, like image data, desired image data must be selected by visually confirming each image data. This requires labor and time and is very difficult if the data amount is enormous.
Hence, data having no text data, like image data, is added with data called metadata and is searched for using the metadata. Additionally, recent digital cameras employ a method of automatically adding metadata called Exif data to a sensed image upon image sensing. Exif data is metadata associated with sensed data, including the shooting date, camera model name, camera owner, shutter speed, aperture value, and shooting location by GPS. It is recent common practice to acquire desired image data by searching for Exif data.
However, it is difficult to accurately select desired image data based on only Exif data used as metadata because its information amount is small. To solve this, in many cases, the user adds, as metadata, information associated with the image data such as an event name “athletic meeting” or “travel” or a personal name afterward, in addition to the Exif data, and searches for the metadata as a search target, thereby selecting desired image data.
However, since much labor and time are necessary for adding metadata to an enormous amount of data one by one, the user may be negligent in adding metadata, resulting in difficulty in selecting desired data.
To solve this problem, Japanese Patent Laid-Open No. 2003-228569 has proposed a technique of extracting keyword group data by morpheme analysis of text data that accompanies data. In the keyword group data extraction technique according to this proposal, however, it is impossible at high probability to use a method of generating metadata by element analysis of text data because many image data have no text data except for Exif data.
Japanese Patent Laid-Open No. 2002-330432 has proposed a technique of generating metadata based on the feature amount of image data. However, the proposed technique may uniformly add the same metadata to similar images and also add the same metadata to an image which is not so relevant to a person.
Japanese Patent Laid-Open No. 2006-099268 has proposed a technique of providing a metadata adding server which requests metadata addition and a requesting function. In this technique, the metadata adding server adds the same metadata to the same person, same event, or same phenomenon, thereby reducing processing of the self terminal and load in adding metadata. However, this technique adds the same metadata to the same person, same event, or same phenomenon. Since the same metadata is added to a number of data, an enormous amount of data are selected by a metadata search, resulting in difficulty in acquiring desired data.
Japanese Patent Laid-Open No. 2006-209516 has proposed a technique of preparing an icon having specific attribute information and storage location information and selecting one or more data and dragging & dropping them, thereby adding the same attribute information and storage location information to the data. This allows the user to simultaneously add the same metadata to a plurality of data to which the same attribute should be added. However, this proposed technique is not suitable for a purpose of adding metadata suitable to each data because metadata prepared in advance is unconditionally added to data selected by the user. If the user selects wrong data, it is impossible to either add appropriate metadata or perform a search by metadata.
Japanese Patent Laid-Open No. 2006-209516 has also disclosed a technique of searching for an element common to the metadata of a plurality of contents. If the metadata have the same constituent elements or constituent elements having a predetermined association, a union of the constituent elements of the metadata is generated. The union is added, as metadata, to the data having same constituent elements or constituent elements having a predetermined association. This makes it possible to add, to each of data having a strong association, the metadata of the respective data as a union. However, as a characteristic feature of this proposed technique, data having an association are added with the same metadata. Even when the user wants to discriminate between data by adding unique metadata to them based on a particularly noticeable difference, the metadata are not unique if the data have any association other than the noticeable point.
Japanese Patent Laid-Open No. 2005-352782 has disclosed a technique of selecting data as a metadata addition target to search for and extract similar data and adding metadata added to the found data to the selected data. This allows, in adding metadata, to add the same metadata as that added previously based on the metadata of the similar data. In this proposed technique, data as a metadata addition target is selected to search for and extract similar data, and metadata added to the found data is added to the selected data. In this case, however, as the metadata adding operation by the user progresses, the possibility that already added metadata is presented for data selected as a metadata addition target increases. If metadata added to data that has happened to be found and extracted as similar data and having little association is presented, the user must do determination every time. When simultaneously adding metadata to a plurality of data to raise the operation efficiency, no search and extraction of highly similar data can be done unless all the data have high similarity.